Open-weight & owned stack · Quantization efficiency
Quantized Qwen 27B Runs On Dual RTX 3060s
A user reported running quantized Qwen3.8-27B on two RTX 3060 GPUs at roughly 44–50 generated tokens per second.
Read the original at reddit.comOpens the publisher's site in a new tab